Despite the early age at onset of both autism and parental concern about child behaviors, the vast majority of children with autism are not identified until after 3-years of age. The overarching goals of Project I are (1) to develop new approaches that will enhance early detection of autism during the second year of life; and (2) to understand the early course of core features of autism, co-occurring symptoms, and temperament in the child in the context of maternal well-being, stress, and resources to inform parent-support and parent-child interventions. Previous longitudinal studies of young children with autism have focused solely on core features of autism, with much less consideration of (a) developmental trajectories of potentially significant co-occurring child features that are not part of the core syndrome; (b) interventions received; or (c) family level factors that may influence, and be influenced by, the early course of children with autism. We plan to recruit a sample of 300 children suspected of autism between the ages of 18- to 30-months and to conduct a five-year, longitudinal study of their early course with respect to: (1) core features of autism, including social functioning, language and communication, and unusual behaviors; (2) co-occurring symptoms, including externalizing, internalizing and regulatory problems as well as low base rate maladaptive behaviors (e.g., head banging); and (3) temperament, using both mother report and observational measures. In addition, we plan to investigate stability and change in (1) parental well-being, (2) parenting stress, and (3) resources from the period of initial diagnosis to school entry. Finally we are proposing to monitor the interventions that children and families receive through this period and to examine the role of (1) treatment intensity (i.e., hours per week) and (2) working alliance (i.e., mother's experience of providers as competent, responsive, and supportive) on both child and family outcomes. A central hypothesis is the presence of early co-occurring symptoms and extreme negative temperamental affect may be early precursors for later diagnosis of co-occurring psychopathology. An outcome of the proposed work will be a first stage autism screener that will include age-specific cut-points for 18- to 36-month old children.